A major barrier to development of disease-modifying therapy for knee osteoarthritis (OA) is the nature of disease progression, a downward spiral involving many tissues. The potential for changing disease course may be highest before this process is in motion. The literature documents cartilage, bone, and meniscal lesions by MRI in knees of some older individuals without radiographic OA, but does not reveal whether they represent early OA. We propose a long-term, longitudinal study to evaluate the clinical significance of such lesions in persons at higher risk for knee OA but without OA in either knee by x-ray at baseline, to advance knowledge of early OA and to help to identify an earlier (than x-ray affords) OA onset window. We have a highly unique opportunity to carry out this study. The Osteoarthritis Initiative (OAI), a study of 4796 individuals with or at higher risk for knee OA, ens follow-up at 96 months. 1244 participants had no x-ray findings of knee OA in either knee at baseline. We propose an additional visit of this subcohort (1226 living) at 120 months to collect data (some not collected after baseline and some not after the 48 month visit) that would enable comprehensive evaluation of clinical definitions of OA at the furthest feasible time point while the OAI clinical sites are still constituted, as well as assessment of MR images being collected at 96 months. The proposed aims are unique, time-sensitive, and could not be completed within the OAI's scope of work. We propose 3 aims in OAI participants in this subcohort: 1) evaluate the association between each lesion - cartilage damage, bone marrow lesion, meniscal tear, meniscal extrusion - by MRI at baseline and risk of incident knee OA by the historic (and current) radiographic criterion by 120 months and by recently proposed MRI criteria by 96 months; 2) evaluate the association between each lesion by MRI at baseline and risk of incident clinical features of knee OA, including ACR criteria, flexion contracture, and persistent knee symptoms by 120 months; and 3) explore the association between each lesion by MRI at baseline and baseline-to-120-month physical function decline by performance and self-report measures of function. Proposed study findings will enable future identification of risk factors for clinically important early lesions and prevention strategies that could be introduced earlier in th lifespan. Ultimately, clinically significant lesions might become a target of disease-modifying intervention at a point when success may be more realizable, before the whole-organ progression spiral.